ConvOptDL

Training

juliap train.jl --batches-per-episode=<INTEGER> --batch-size=<INTEGER> --num-episodes=<INTEGER> --log-period=<INTEGER> <DATA_FILE.jls>

ex: juliap train.jl --batches-per-episode=400 --batch-size=20 --num-episodes=20 --log-period=20 CIFAR_FS-train.jls

GPU support
implement Block LU factorization in the website below for better performance
- https://locuslab.github.io/qpth/
better QP solver: primal-dual interior point method
- As mentioned in Mattingley & Boyd (2012)

Name		Name	Last commit message	Last commit date
Latest commit History 38 Commits
.github/workflows		.github/workflows
.vscode		.vscode
slides		slides
src		src
test		test
.gitignore		.gitignore
LICENSE		LICENSE
Project.toml		Project.toml
README.md		README.md
test.jl		test.jl
train.jl		train.jl